Latest News - Page 125

NVIDIA unveiled DLSS 5 at GTC 2026, and the new technology has been met with immense pushback from gamers, who are concerned their video games will be poisoned with "AI slop" currently taking over numerous social media platforms.

DLSS 5 is on its way to release sometime in Fall 2026, and during the announcement, NVIDIA's CEO Jensen Huang described this next generation of DLSS as the "GPT moment for graphics," a reference to the release of the first AI model. In its announcement of DLSS 5, NVIDIA states that the next-generation AI upscaler will be able to "bridge the divide between rendering and reality" by analyzing light sources, characters, hair, shadows, skin, and other aspects of a single frame.

Then, DLSS 5 processes real-time game data and produces an image that, in some cases, such as the one NVIDIA used to demonstrate the technology, can change an image's appearance almost entirely. The discrepancy between the original image and the result when DLSS 5 is turned on is evident in many of the examples NVIDIA showed, producing a result that is stunningly different, which has sparked a steadfast, defiant stance by gamers against the technology's adoption.

Rumors of CD Projekt Red quietly working on a new Witcher 3 DLC have heated up once again, which is great for any fans of the iconic RPG that were a little dismayed at the idea CD Projekt Red was working on a new Witcher 3 DLC as the developer just confirmed it isn't working on something "secret" for Cyberpunk 2077. The thought being, "If they aren't working on something secret for Cyberpunk 2077, why would they be working on something for The Witcher 3: Wild Hunt".

The rumors of the new DLC for Wild Hunt pointed to CD Projekt Red expanding on the region Velen, and the DLC would focus on bridging the gap between the events at the end of The Witcher 3 and the beginning of The Witcher 4. The idea generally sat well with many Witcher 3 fans, including myself, as what better way to get excited for The Witcher 4 than to dive back into the incredible world of The Witcher 3 to do a short-story-length piece of content that leads you right into the new game?

It would get long-time Witcher 3 fans opening up one of their favorite games again, reigniting their love for it, and then, that translates into purchase of The Witcher 4, or, at least that's what CD Projekt Red would be hoping. Now, a Polish analyst has penned a new report that claims they anticipate CD Projekt Red will release this rumored DLC sometime in September 2026, as they want to get it out before the release of Rockstar's Grand Theft Auto VI, which is teed up to be one of, if not the biggest launch in gaming history.

NVIDIA has unveiled the new Vera Rubin AI platform, specifically designed to advance AI into its next phase of reasoning, automation, agentic-based workloads, and continuous responses. However, not all that power will be allocated to the numerous AI server farms around the world; some of it is destined for off-planet use.

During NVIDIA's GTC 2026 presentation, company CEO Jensen Huang unveiled the Vera Rubin Space Module, which Huang says will deliver up to 25x times more AI compute than Hopper-generation H100. According to the announcement, six commercial space companies have already adopted the new Vera Rubin Space Module, which NVIDIA says is specifically designed for orbital data centers that are running LLMs in space.

Huang explained during the keynote that NVIDIA is seeking a way to expand its AI computing stack beyond Earth, leading to the creation of "AI factories in space". The Vera Rubin Space Module is designed to solve a current problem in data transfer in space: satellites capture large swaths of data and then have to send it back to Earth for processing. That process introduces latency and bandwidth requirements, but with the Vera Rubin Space Module, the idea is that inference compute is much closer to where the data is being generated, reducing latency, processing time, etc.

In a new report in Taiwan's United Daily News publication (auto-translated), MSI's General Manager Huang Jinqing has warned investors in a recent briefing that 2026 is shaping up to be the "most severe year since the company was founded." Naturally, this is due to current DRAM, storage, and even GPU supply and pricing issues driven by demand for all things AI and AI infrastructure.

And when it comes to graphics cards, of which MSI is a key GeForce RTX partner, the company estimates a supply gap of around 20%, meaning NVIDIA isn't able to provide the stock and GPUs needed to maintain the broader consumer PC market. Not only that, but MSI also predicts this will lead to the PC market shrinking by 10 to 20% in 2026, a larger figure than the one we recently reported from the International Data Corporation (IDC).

When it comes to the company's plans to weather the storm, so to speak, in 2026, it will focus on the mid-range and high-end GeForce RTX market, with cards like the GeForce RTX 5060 and RTX 5070, and the price of its graphics cards will increase by anywhere between 15 and 30%. The goal is to increase profit and profit margins as the overall volume decreases.

At NVIDIA GTC 2026 in San Jose, California, Micron has announced that new AI-optimized memory and storage technologies for the upcoming NVIDIA Vera Rubin platform and other AI systems are now in "high-volume" production. This includes HBM4 36GB 12H memory, which Micron says achieves over 11 Gb/s per pin, opening the door to over 2.8 TB/s of memory bandwidth.

According to the company, this is a 2.3X increase in memory bandwidth and a 20% improvement in power efficiency over HBM3E, which is still one of the fastest, highest-capacity high-bandwidth memory solutions for AI available today. And with advanced packaging, Micron is already shipping HBM4 48GB 16H samples to customers.

HBM4 is built for NVIDIA Vera Rubin and the immediate future of cutting-edge AI, but Micron has also announced solutions designed for NVIDIA Vera Rubin NVL72 and standalone NVIDIA Vera CPU systems with Micron SOCAMM2, offering up to 2TB of memory and 1.2 TB/s of bandwidth per CPU, with capacities ranging from 48GB to 256GB.

KIOXIA's open-source AiSAQ (All-in-Storage ANNS with Product Quantization) technology is something we've covered in the past, and it has been a game-changer for AI workloads with Retrieval-Augmented Generation (RAG) pipelines by offloading vectorized data from costly DRAM to a more efficient, cost-effective SSD solution.

This opens the door to scaling, especially for running more complex AI workloads. And when it comes to large-scale workloads and systems, at NVIDIA GTC 2026, KIOXIA demonstrated that, using the NVIDIA cuVS Library with KIOXIA AiSAQ Technology, it successfully scaled high-dimensional vector search to an impressive 4.8 billion vectors on a single server.

With the NVIDIA cuVS Library offering a "significant reduction in index build time by leveraging GPU acceleration," this breakthrough is a big step forward for retrieval augmented generation (RAG) search solutions. And with that, KIOXIA is already looking toward supporting larger-scale event deployments that go beyond 4.8 billion vectors.

At NVIDIA GTC 2026 this week, the annual all-things-AI event, KIOXIA announced it is developing a Super High IOPS SSD, which it described as a "new type of SSD" built for AI systems of today and the future. What makes it new is that it allows the GPU to access the SSD's high-speed flash memory as an extension of existing High Bandwidth Memory (HBM) in AI systems.

Essentially, with KIOXIA's new Super High IOPS SSD, AI systems will be able to dramatically increase the available memory capacity with the help of high-speed SSDs. Part of the KIOXIA GP Series, the company confirms that development is underway and that evaluation samples will be sent out to select customers by the end of the year.

DRAM capacity, or memory size, is one of the limiting factors and the highest costs in scaling large AI systems. And with that, this new SSD is KIOXIA's contribution to the NVIDIA Storage-Next initiative, which is about addressing this shift by incorporating additional solutions for data-intensive AI workloads constrained by DRAM capacity or HBM availability.

NVIDIA's CEO Jensen Huang has officially unveiled its next-generation AI platform, which includes seven chips and six individual racks.

Vera Rubin was unveiled at GTC 2026, where Huang broke down the components that make up NVIDIA's new flagship AI platform. NVIDIA has positioned the unveiling of Vera Rubin as the foundation of a new era of AI infrastructure, with the platform being specifically designed around AI inference and agent-based workloads. At its core, Vera Rubin is a full-stack AI supercomputing platform that combines a set of tightly integrated components, including the Vera CPU, Rubin GPU, NVLink interconnect, networking, and data processing units.

Huang explained during the GTC Keynote that AI is moving toward continuous generation of responses, decision-making, and action-taking. The NVIDIA CEO added that Vera Rubin has been specifically designed to support this new direction of AI inference. Notably, NVIDIA's Vera Rubin is an AI system designed to reason, plan, and act autonomously. As for performance, NVIDIA has said Vera Rubin will cut inference token costs by up to 10x and reduce the number of GPUs required for complex models.

DLSS 5 has been announced, and NVIDIA has confirmed it's on track for a Fall 2026 release. With NVIDIA CEO Jensen Huang describing it as a "GPT moment for graphics," it takes in-game data to leverage neural rendering, delivering dramatic improvements in lighting, visual fidelity, and both character and environment detail in real-time.

"Bridging the divide between rendering and reality, DLSS 5 empowers game developers to deliver a new level of photoreal computer graphics previously only achieved in Hollywood visual effects," NVIDIA writes in the announcement. DLSS 5 utilizes a powerful new AI model trained on game objects and graphics, then adds per-pixel effects like realistic path-traced lighting, materials, detailed hair, and 'subsurface scattering on skin' to add an impressive amount of detail.

Developers also have control over the intensity, color grading, and other elements of DLSS 5 to ensure that the art direction, aesthetics, and visual detail are consistent with the game's artistic intent. That said, the announcement has been controversial, with online backlash claiming that some of the results look like AI-generated photo filters and video effects. Either way, it's hard to deny the staggering difference DLSS 5 can make, and with that, here's the list of the first 16 games getting DLSS 5 support.

NVIDIA's surprise DLSS 5 announcement at GTC 2026 has been controversial, with several comments from the wider gaming community and media comparing it to an Instagram-like AI filter that beautifies character models. This reaction mostly stems from one key example showcasing the character Grace in Capcom's Resident Evil Requiem, where there's the impression that DLSS 5 is not only changing the character's look but also adding additional details like makeup.

At its core, DLSS 5 is a game-changer and the first look at the future of in-game visuals, aiming to deliver lifelike, photorealistic lighting. DLSS 5 is powered by a new, powerful AI model (the demos we see are running on two GeForce RTX 5090s) that takes game data, models, geometry, textures, lighting, and more to render a photorealistic frame.

And like other DLSS technologies that use game and frame data, this will apparently ensure that the result is consistent from frame to frame and across different play sessions. In addition, developers have control over the intensity of the effect, color grading, and masking (i.e., choosing which elements in a scene not to apply DLSS 5 to), with NVIDIA confirming that the Resident Evil Requiem footage shown was tuned and approved by Capcom.

"Twenty-five years after NVIDIA invented the programmable shader, we are reinventing computer graphics once again," said Jensen Huang, founder and CEO of NVIDIA at GTC. "DLSS 5 is the GPT moment for graphics, blending hand-crafted rendering with generative AI to deliver a dramatic leap in visual realism while preserving the control artists need for creative expression."

NVIDIA's surprise DLSS 5 announcement this week has garnered a wide range of responses from the wider gaming community and media, and it's not hard to see why. The announcement video above, presenting DLSS 5 Off and DLSS 5 On examples in games like Resident Evil Requiem, Hogwarts Legacy, EA SPORTS FC, and Starfield, is jaw-dropping for the difference it makes in lighting, realism, and even detail. Mind-blowing even.

As NVIDIA briefly summarizes, DLSS 5 is a "real-time neural rendering model that infuses pixels with photoreal lighting and materials," taking a game's color and motion vectors as input. From there, developers have access to detailed controls over DLSS 5's intensity, color grading, masking, and more, to "maintain each game's unique aesthetic." As impressively game-changing as it is, and even though DLSS 5 doesn't change things like character models and animation, some of the results can look a little jarring in that Uncanny Valley sort of way.

WB Games Montreal has been hit with layoffs as its parent company prepares to merge with Paramount.

WB Games' Montreal-based studio has made a few layoffs across various fields, affecting devs that have been with the studio for years--including one that has 18 years with the group.

Over on LinkedIn, at least three developers have revealed that they have been let go from positions at WB Games Montreal. This follows a swath of layoffs conducted 8 months or so ago, which seems to be the results of WB management's earlier decisions to scale back its gaming ambitions and focus on four key franchises--Harry Potter, Mortal Kombat, DC, and Game of Thrones. This strategy was led by the prospect of cost savings and reducing spending on higher-risk projects.

After years of waiting, Google is finally introducing native support for ARM64 devices with Linux. This support comes several years after Chrome added support for ARM64 devices running macOS and Windows. According to a post at the Chromium blog, Chrome for ARM64 Linux devices is currently due in Q2 of 2026.

Linux users with ARM64 processors have been able to use the open-source Chromium browser for some time, but while its core browser functionality works well enough, users should appreciate being able to access the full stack of Chrome services via a Google account login.

It's a little surprising that official support has taken so long to implement. Linux on ARM servers has long been a stable market, and Chromebooks with the full Google product stack have ARM support, as does Android. So, gluing the two together for Linux end users should not be too much of a technical challenge.

Saying one thing is a million times more something than another thing is often hyperbole, but during a recent GDC 2026 presentation, NVIDIA's John Spitzer said exactly that when it comes to path tracing performance on future GeForce RTX graphics cards. Of course, this comes with a caveat: performance compared to NVIDIA's pre-RTX Pascal-era GeForce GTX graphics cards that lacked dedicated ray-tracing and AI hardware. Yes, it's all thanks to AI.

Real-time path tracing or full ray tracing is so demanding on GPU hardware that it's only possible thanks to a wide range of AI-powered rendering technologies such as DLSS Super Resolution, Frame Generation, RTX Mega Geometry, and more. As these features are available on the current RTX Blackwell-powered GeForce RTX 50 Series, NVIDIA says it has already achieved 10,000X faster path-tracing performance than in the Pascal era.

"If you look at the performance there with just a software RT core to today, where we have fourth-generation RT cores, we have third-generation Tensor cores, we have DLSS 4.5, which is able to infer 23 out of 24 pixels rendered," NVIDIA VP of Developer & Performance Technology, John Spitzer, said. "These are multiplicative, that you can multiply them all together to get a scaling factor that, combined with the algorithm, eventually gave a 10,000 times that we've improved the performance over the last 10 years."

COLORFUL has announced a new line-up of GeForce RTX 50 Series graphics cards, powered by the latest RTX Blackwell architecture. The new iGame GeForce RTX 50 Ultra Series blends a similar hip-hop-inspired design and aesthetic to last year's Ultra lineup, albeit in a darker black colorway that highlights the purple flourishes.

Visually, they look impressive, with a gradient of purples and neons in the holographic finish and RGB graffiti on the side. With a large heatsink and triple fans for cooling, this new lineup covers all models from the GeForce RTX 5060 up to the GeForce RTX 5070 Ti. All options ship with an out-of-the-box OC Mode, with the most powerful model including an iGame GeForce RTX 5070 Ti Ultra OC SFF variant for compact builds.

In addition, COLORFUL has announced the new iGame GeForce RTX 50 Ultra DUO Series, featuring a similarly stylish design with compact dual-fan cooling. COLORFUL is releasing iGame GeForce RTX 5060 Ultra DUO and iGame GeForce RTX 5060 Ti Ultra DUO models, including RTX 5060 Ti variants with 8GB or 16GB of VRAM.

Silicon Motion has announced the launch of its latest PCIe Gen5 x4 NVMe SSD controller, the SM8008. Designed and purpose-built for data center drives and enterprise storage, where NVMe SSDs are becoming the go-to option for the AI era, the SM8008 is all about delivering performance, power efficiency, and stability.

Built on TSMC's 6nm process technology, the SM8008 delivers up to 14 GB/s or 14,000 MB/sec with over 2.3 million random IOPS (4K) while maintaining a power draw of under 5W. NVMe 2.0a and OCP Hyperscale NVMe Boot SSD Specification Version 1.0 compliant, you've also got inline DDR4-3200 or LPDDR4-3200 to enhance power efficiency and affordability for large-scale deployments.

On that note, Silicon Motion says that SSDs with the SM8008 controller can operate continuously across thousands or even "millions" of servers. Although it's difficult to visualize what millions of SSDs in a data center would look like, the point is that with PCIe Gen5 speeds and power consumption of under 5W, this translates to "significant reductions in total data center power consumption."

Playing a modern PC game, we have all grown accustomed to the 'shader compilation' process, which can take seconds or several minutes to complete. This process, designed to minimize stutter that occurs when new shaders are compiled in real time, also increases the time it takes to get into the action. And it's not always a one-and-done thing, as changes to settings and game updates can restart the shader compilation process.

At GDC 2026, Microsoft announced that it's bringing Advanced Shader Delivery for Windows to PC gamers, with NVIDIA confirming that it will be available for GeForce RTX gamers later this year. It's an ambitious update that is rolling out via the latest AgilitySDK 1.619 release, where developers can generate a "state object database (SODB) file" and then use an "offline compiler to compile the state objects into a precompiled shader database (PSDB)."

Although that's the technical summary of what's involved, the good news for gamers is that pre-compiled shaders for their specific GPU model will be downloaded as part of installing the full game, so the process will be automated and seamless. And for developers, there's also a new Stats API that can be used for optimization and seeing how the game runs in relation to the "shader cache hit rate."

Thermal Grizzly's WireView Pro II is designed not only to measure the power consumption of modern graphics cards or GPUs that use the new 12V-2x6 connector, like NVIDIA's flagship GeForce RTX 5090, but also to prevent potential damage. With per-pin current monitoring, it can detect and alert users of any potential load issues.

The WireView Pro II is an enhanced version of the company's original design and features a CNC-machined aluminum housing with active cooling via a 30-mm fan and 90-degree cable routing. It sports a 320x170-pixel TFT-IPS color display for real-time monitoring of critical measurements such as total power, highest-current pin, voltage, and hottest sensor. And with safety being a key feature of the WireView Pro II, there are both audible and visual warnings if limits are exceeded.

Plus, there's automatic data logging via the included USB-C-to-USB-2.0 cable, making it a fantastic tool for overclockers and those with high-end systems looking to keep tabs on GPU power usage. And with the rise of white-colored PC builds and components, Thermal Grizzly has announced it's launching a WireView Pro II White Edition, available to pre-order via the company's official storefront.

Marathon is out, and pulling respectable concurrent numbers on Steam while holding meaningful positions on platform sales charts...but the same can't be said about Destiny 2.

Forbes' Paul Tassi has noticed a trend with Destiny 2's numbers: Bungie's nearly 10-year-old live shooter has lost a lot of its players. How much is a lot? About 90% since the last expansion, Edge of Fate, which was deemed as one of the game's all-time worst updates.

SteamDB's sales charts for the lifetime of the game also revealed some other disappointing news: Destiny 2 is apparently at its lowest point, player-wise, on Steam. The game currently has less than 8,000 users on Steam, and according to the SteamDB, March 2026 represents the lowest point of all time for Destiny 2's player base.

Across government institutions, universities, research centers, and corporate environments, the volume of paper-based records continues to grow. Contracts, research documents, administrative files, and historical records often remain stored in physical archives that are difficult to manage, search, or protect over time. As organizations accelerate their digital transformation strategies, archive digitization has become a critical step toward improving accessibility, efficiency, and long-term information preservation.

Traditional scanning methods once focused mainly on converting paper documents into image files. However, modern archive digitization now requires more advanced workflows that combine high-speed scanning hardware with intelligent software capable of organizing, processing, and managing digital records. Integrated scanning solutions are helping organizations move beyond simple document capture and toward smarter archive management systems.

Large-scale archive projects present several challenges. Physical records may include fragile documents, bound books, or historical materials that cannot be easily flattened for scanning on conventional scanners. At the same time, organizations handling thousands of pages daily require systems that can process documents quickly without sacrificing accuracy or image quality.